All ETDs from UAB

Advisory Committee Chair

Nasim Uddin

Advisory Committee Members

Christopher Waldron

Jason Kirby

Mohamed Hider

Document Type

Thesis

Date of Award

2015

Degree Name by School

Master of Science in Civil Engineering (MSCE) School of Engineering

Abstract

Bridge structures are subjected to continuous degradation due to environmental effects and increase in traffic loading. Bridge monitoring is a key element of any maintenance strategy as it can provide early signs of structural deterioration. In recent years, the demand towards sensor-based bridge monitoring in lieu of visual inspection, has increased. However, sensor-based monitoring is costly and requires significant maintenance. Recent studies on bridge damage monitoring have focused on the instrumentation of a passing vehicle instead of the bridge. This approach is referred to ‘drive-by’ bridge inspection (Kim and Kawatani, 2009). Using this approach, the bridge itself is not instrumented, which makes the concept so far cost effective comparing to the traditional health monitoring methods. This thesis theoretically assesses the ability of using an instrumented vehicle to detect structural damage in bridges. This thesis focuses on three main objectives. First, the feasibility to use two different indicators (''Power Spectral Density'' and ''Apparent Profiles'') to detect bridge damage is explored. For ‘Apparent Profile’ term, when a vehicle passes over a certain profile without a bridge, it is dynamically excited. Using its acceleration data as an input, it is possible to ‘back-calculate’ the profile that causes the excitation. If the vehicle starts to pass over the same profile with a structurally healthy bridge, the back-calculated profile will be contaminated by bridge displacement which is what referred to as the ‘Apparent Profile’. In summary, the ‘Apparent Profile’ refers to thesummation of a ‘Road Profile’ and ‘Bridge Displacement’. As you repeat this process for the same road profile with a damaged bridge, it results in a ‘Damaged Apparent Profile’. Subtracting the ‘Damaged Apparent Profile’ from the ‘Undamaged Apparent Profile’ provides an indirect indication of the damage level in the bridge.The second objective of this thesis is to use the Inverse Dynamics Optimization Technique in ‘back-calculating’ the ‘Apparent Profile’. This technique first emerged in the late 1990's (Law et al., 1997) to be used for B-WIM systems (Bridge-Weight In Motion). In this study, the technique is used to calculate the dynamic response of the vehicle model. The inverse Dynamics Optimization is an inverse solution for the partial differential equation (which here is the equation of motion for the vehicle), where the right hand side (the applied force) is calculated using only a measurement from the left hand side (the vehicle acceleration, velocity or displacement). The Inverse Dynamics Technique has two main contributions in ‘back-calculating’ the ‘Apparent Profile’. First, it gives the full dynamic response for the vehicle using only one measurement and therefore only axle acceleration or velocity can be used to ‘back-calculate’ the apparent profile, instead of measuring the body and axle mass acceleration data. Second, it is not an ill-conditioned solution, where the solution is not affected by the signal noise or the poor formulates stiffness/mass matrices. The final objective in this study is to use the ‘Apparent Profile’ to determine an unknown bridge damage location and its value. The Inverse Dynamics Optimization Algorithm is then used in the back-calculation of the Apparent Profile.

Included in

Engineering Commons

Share

COinS